Improving diagnostic accuracy of bacterial pharyngitis by near patient measurement of C-reactive protein (CRP)

Abstract

BACKGROUND: Sore throat or pharyngitis is an extremely prevalent condition in primary care. There is a diagnostic dilemma in differentiating bacterial and non-bacterial infections for adequate use of antibiotics. Standard diagnostic procedures take too long for an immediate decision. AIM: To evaluate, if near patient C-reactive protein measurement in the general practice surgery improves diagnostic accuracy. METHOD: One hundred and seventy-nine consecutive patients with sore throat, from 15 general practitioners (GPs) in southern Germany (phase 1) and 161 consecutive patients from 14 GPs (phase 2), were examined physically and a throat-swab was taken and white blood-cell count (WBC) and CRP-measurement were performed. In phase 1, CRP was measured centrally to assess the method's diagnostic value and the adequate threshold. In the second phase, near patient CRP was measured and CRP values were used to make a diagnosis. RESULTS: Using relative operating characteristics (ROC) analysis, the diagnostic value of CRP measurement was much better than WBC count (area under curve = 0.85 versus 0.68). All diagnostic parameters improved when using the near patient CRP measurement. Sensitivity went up from 0.61 (95% confidence interval = 0.45-0.75) to 0.78 (0.61-0.90), specificity went up from 0.73 (0.65-0.81) to 0.82 (0.73-0.88). Positive and negative predictive value improved significantly as well. Diagnostic accuracy went up from 70.1% to 81.0%. Out of 1000 theoretical patients with sore throat, 109 more will be treated correctly when using CRP measurement as a diagnostic tool. CONCLUSIONS: Use of near patient CRP measurement can improve diagnostic accuracy in the differentiation of bacterial and non-bacterial pharyngitis in primary care, and potentially results in a more adequate use of antibiotics.